交通视频图像中运动目标的提取  

Extraction of the Moving Object in the Traffic Video

作　　者：张虹波[1] 匡银虎[1] 

机构地区：[1]宁夏大学物电气信息学院,银川750021

出　　处：《现代计算机（中旬刊）》2014年第10期61-65,共5页Modern Computer

基　　金：2014宁夏高等学校科学技术研究项目资助

摘　　要：主要阐述交通视频检测中背景模型的建立,在普通均值法提取背景的基础上,提出改进的均值提取法,并通过实验验证和对比,论证改进的均值法在提取背景图像方面,可以减少背景中的车辆幻影,得到效果比较好的背景图,进而对视频中运动目标的提取作介绍,并通过实验得到所需的前景图,即运动目标,为后续的车流量计数工作做好基础准备。Mainly describes the establishment of background traffic video detection model, extraction in general average method on the basis of the background, puts forward the improved mean extraction method, and through experimental verification and comparison, demonstrates the advanced mean value method in extracting the background image, can reduce vehicle phantom background, get a better background map. And introduces the extraction of moving objects in video, and the prospect of graph needed to get through the experiment, namely the moving object, to the foundation to prepare for the follow-up work flow counting cars.

关 键 词：视频检测技术 智能交通 背景差分 虚拟线圈 交通车流量 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]